Abstract
PURPOSE
To evaluate outcomes of choroidal melanoma patients treated with 125I or 103Pd plaque brachytherapy.
METHODS AND MATERIALS
From 1993 to 2012, our institution treated 160 patients with 103Pd (56.1%) and 125 patients with 125I (43.9%) plaque brachytherapy. Tumor outcomes, visual acuity (VA), and toxicity were compared. Multivariate analyses (MVAs) and propensity score analysis were used to help address differences in baseline characteristics.
RESULTS
Median followup was longer for 125I patients, 52.7 vs. 43.5 months (p < 0.01). At baseline, 103Pd patients had lower rates of VA worse than 20/200 (4.4% vs. 16%, p = 0.002), T3–T4 tumors (17.5% vs. 32.8%, p = 0.03), and transpupillary thermotherapy use (3.1% vs. 9.6%, p = 0.001). Both 103Pd and 125I provided >90% 3-year overall survival and >93% 5-year secondary enucleation-free survival. On MVA, radionuclide was not predictive for tumor outcomes. A higher percentage maintained vision better than 20/40 with 103Pd (63% vs. 35%, p = 0.007) at 3 years. MVA demonstrated 103Pd radionuclide (odds ratio [OR]: 2.12, p = 0.028) and tumor height ≤5 mm (OR: 2.78, p = 0.017) were associated with VA better than 20/40. Propensity score analysis matched 23 125I with 107 103Pd patients. 103Pd continued to predict better VA at 3 years (OR: 8.10, p = 0.014). On MVA for the development of VA worse than 20/200 or degree of vision loss, radionuclide was not significant. Lower rates of radiation retinopathy were seen with 103Pd than 125I (3 years: 47.3% vs. 63.9%, p = 0.016), with radionuclide significant in MVA.
CONCLUSIONS
Both 125I and 103Pd achieve excellent tumor control. An increased probability of long-term VA better than 20/40 and reduced risk of radiation retinopathy is associated with 103Pd.
Keywords: Choroidal melanoma, Episcleral plaque brachytherapy, Palladium, Iodine
Introduction
Choroidal melanoma is the most common primary intraocular malignancy (1). For many years, enucleation was the standard therapy for patients with choroidal melanoma. From 1986 to 2003, the Collaborative Ocular Melanoma Study (COMS) group randomized patients with medium-sized tumors to enucleation or 125I episcleral plaque brachytherapy (EPB). These studies demonstrated no difference in overall survival (OS) (2) and distant metastases (3). EPB, however, had more favorable visual acuity (VA) outcomes: 49% of these patients with baseline vision better than 20/40 retained VA of 20/40 or better at 3 years (4). Based on these results, the American Brachytherapy Society has recommended 125I EPB as one alternative to enucleation for the treatment of medium size choroidal melanoma.
Although 125I has demonstrated >85% 5-year local tumor control rates, 41–58% of patients develop significant deterioration in vision (vision > 20/200) in part due to radiation complications, including radiation retinopathy and vitreous hemorrhage (5, 6). 103Pd is a lower energy gamma-emitting radionuclide with a shallower depth of penetration than 125I (7); with this translating into less potential dose delivered to normal ocular structures (8–15), 103Pd may decrease related radiation toxicity. Consistent with this hypothesis, Finger et al. first demonstrated in 2002 that 82% of patients (n = 68) treated with 103Pd EPB had maintained VA of 20/200 or better at 3 years (16).
Based on these considerations, our institution began treating nearly all cases of choroidal melanoma with 103Pd in 2004. As the COMS clinical trial excluded patients with tumors within 2 mm of the optic disc, 125I was used for patients with tumors in contact with the optic nerve or those with very poor initial VA. 125I—which is less expensive than 103Pd—was also used in situations where cost was prohibitive. With no studies to our knowledge directly comparing patient outcomes after treatment with 103Pd to 125I plaque brachytherapy, the purpose of this study was to compare disease outcomes, VA, and toxicity of patients treated with these radionuclides at our institution.
Methods and materials
Patients
Following Institutional Review Board approval, we reviewed the charts of all patients treated with radionuclide eye plaques between 1993 and 2012. The following exclusion criteria were applied: patients with incomplete demographic, pretreatment VA or treatment data, iris or ciliary body melanoma, and patients with metastatic disease at initial presentation.
Treatment methods
For 125I, ABS guidelines recommended prescribing 85 Gy to the apex, including for lesions less than 5 mm (5). The COMS study prescribed 85 Gy to 5 mm for lesion less than 5 mm in height; for tumors greater than 5 mm in height, the dose was prescribed to the tumor apex (2). At our institution, we followed the COMS approach. Plaque size (12) was chosen based on tumor size, allowing a 2-mm margin around the circumference of the tumor. For tumors close to the optic nerve, a notched plaque was chosen. Plaques were placed by oncologic ophthalmologist (authors TAM, TA, and CB) and were removed after 7 days. In an effort to maximize tumor control, transpupillary thermotherapy (TTT) was applied before or during iodine (n = 33) and palladium (n = 11) plaque brachytherapy. There was no selection bias relating to adding TTT to either radionuclide.
Statistics
Baseline patient data included age at time of treatment, gender, presence or absence of diabetes mellitus, best corrected VA before treatment, T stage (based on AJCC seventh edition staging manual), tumor area (length × width), height, and location (equator, posterior, equator/posterior, equator/anterior) and presence of retinal detachment. Treatment data included date of plaque placement, radionuclide used (125I or 103Pd), plaque size, plaque shape (notched vs. nonnotched), radiation dose, and prescription depth. Distance of the tumor from the optic nerve and fovea were also noted; these values are reported relative to the size of the globe to account for variations in cameras and images throughout the 20-year time period.
Followup data included length of followup, VA at annual followup visits, the date of diagnosis of complications of radiotherapy (radiation retinopathy or vitreous hemorrhage), distant metastases, secondary enucleation, and date of death from all causes of mortality.
Baseline and followup VA measurements were recorded using a Snellen VA scale. Count fingers VA was converted to a Snellen VA recording (17). Hand motion and light perception vision were recorded as a decimal acuity of 0.005, whereas no light perception and enucleation were given a decimal acuity of one line worse than the worst VA (0.0015) (18).
To determine the effect of radionuclides on VA, three VA measures were used: preservation, loss, and the degree of change. VA preservation was defined as posttreatment vision better than 20/40. VA loss was defined as impairment in vision to 20/200 or worse. Degree of VA loss was recorded as (1) improved, stable, (2) decrease in up to 1–2 lines on the Snellen chart (e.g., 20/30 to 20/40 = 1 line lost), (3) 3–5 lines, or (4) ≥6 lines. Patient characteristics were compared between those treated with 125I and 103Pd by two sample t-test for continuous variables and chi-square tests or Fisher’s exact tests for categorical variables, where appropriate. VA preservation and loss were analyzed as dichotomous variables; rates were compared at 1 and 3 years utilizing chi-squared or Fisher’s exact test. Time-to-event analyses were measured from the date of brachytherapy. Enucleation-free survival, distant metastases-free survival, and OS were estimated by the Kaplan–Meier product-limit method, and the log-rank test was used to assess the differences in distribution between patients treated with 125I and 103Pd. For radiation retinopathy analyses, patients were censored for salvage enucleation. Covariates found to be significant or close to significant (p < 0.20) on univariate analysis were entered into the multivariate analysis (MVA), with radionuclide entered in last.
Additionally, we compared the iodine and palladium groups using a propensity score analysis. Patients with missing any clinical characteristic, baseline vision worse than 20/200, with a tumor height greater than 10 mm, or with TTT prior or concurrently were removed from this additional analysis. An inverse probability treatment weighting propensity score analysis was performed, where observations were reweighted based on a function of their propensity score. The propensity scores were derived using the following variables to predict treatment group: base diameter, T stage, age, tumor height, gender, receiving TTT, distance from tumor to optic nerve, distance from tumor to fovea, and baseline vision.
All analyses were carried out using the SAS 9.3 version statistical software package (SAS Inst., Cary, NC). All statistical analyses were two-sided, and p-values <0.05 were considered statistically significant.
Results
Patient population, tumor, and treatment characteristics
Between 1993 and 2012, 285 patients with choroidal melanoma were treated with plaque brachytherapy. One hundred twenty-five patients were treated with 125I (43.9%); 160 patients were treated 103Pd (56.1%), with 103Pd treatment commencing in 2004. Median followup time was 51.9 months (range 7.1–163.7) in the 125I group and 43.5 months (range 6.8–119.8) in the 103Pd group (p < 0.01). Table 1 summarizes the patients, disease, and treatment characteristics of each cohort. The 125I cohort had larger tumors, as measured by larger median base area (136.6 mm2 vs. 105.9 mm2, p < 0.01), larger median height (4.6 mm vs. 2.9 mm, p < 0.01), and more advanced T stage (T3–T4, 32.8% vs. 17.5%, p = 0.03); furthermore, the 125I group had more frequent TTT treatment before or with brachytherapy (9.6% vs. 3.1%, p = 0.005) and higher rate of baseline VA worse than 20/200 (16.0% vs. 4.4%, p = 0.002); VA better than 20/40 was similar (p = 0.11) between iodine (42.4%) and palladium (51.9%). Baseline retinal detachment was more common in the palladium cohort (32.5% vs. 17.6%, p = 0.004). Relative distance of the tumor to the fovea was similar between cohorts. Although close proximity to the optic nerve (i.e., 0–1 relative units) was also similar (18.4% vs. 15.0%), there were a greater number of unknown distance values in the 125I cohort (29.6% vs. 1.3%, p = 0.046). The median prescribed dose was 85.3 Gy for 125I and 84.6 for 103Pd; the median prescription point was 5 mm for both radionuclides.
Table 1.
Baseline patient characteristics
| Covariates | 125I (n = 125) | 103Pd (n = 160) | p-value |
|---|---|---|---|
| Gender (%) | |||
| Male | 72 (57.6) | 81 (50.6) | 0.24 |
| Female | 53 (42.4) | 79 (49.4) | |
| Median age (y) | 64 (range: 16–93) | 64 (range: 23–87) | 0.96 |
| Diabetes (%) | 0.73 | ||
| Yes | 18 (14.5) | 20 (13.1) | |
| No | 107 (85.5) | 140 (86.9) | |
| Laterality (%) | |||
| OS | 65 (52.0) | 76 (47.5) | 0.45 |
| OD | 60 (48.0) | 84 (52.5) | |
| Visual acuity (20/X) (%) | |||
| 20 | 24 (19.2) | 37 (23.1) | 0.15 |
| >20–<50 | 47 (37.6) | 68 (42.5) | |
| ≥50–<100 | 20 (16.0) | 35 (21.9) | |
| ≥100–500 | 23 (18.4) | 16 (10.0) | |
| ≥500 | 5 (4.0) | 2 (1.3) | |
| CF | 3 (2.4) | 0 (0) | |
| HM | 2 (1.6) | 2 (1.3) | |
| LP | 1 (0.8) | 0 (0) | |
| Visual acuity (%) | |||
| ≤20/200 | 105 (84.0) | 153 (95.6) | <0.01 |
| >20/200 | 20 (16.0) | 7 (4.4) | |
| Location (clock position) (%) | |||
| 12:00–3:00 | 38 (30.4) | 42 (26.3) | 0.43 |
| 3:30–6:00 | 33 (26.4) | 36 (22.5) | |
| 6:30–9:00 | 31 (24.8) | 54 (33.8) | |
| 9:30–11:30 | 23 (18.4) | 28 (17.5) | |
| Location (globe position) (%) | |||
| Equator | 30 (30.4) | 49 (30.6) | 0.28 |
| Posterior | 17 (13.6) | 18 (11.3) | |
| Equator and posterior | 47 (37.6) | 76 (47.5) | |
| Equator and anterior | 21 (16.8) | 15 (9.4) | |
| Ora Serrata | 2 (1.6) | 2 (1.3) | |
| Median base area (mm2) | 136.6 (range: 13.5–312.8) | 105.9 (range: 0.1–1075.8) | <0.01 |
| Median height (mm) | 4.6 (range: 1.1–13.5) | 2.9 (range: 0.7–10.3) | <0.01 |
| T stage (%) | |||
| 1 | 9 (7.2) | 17 (10.6) | 0.03 |
| 2 | 75 (60.0) | 115 (71.9) | |
| 3 | 39 (31.2) | 27 (16.9) | |
| 4 | 2 (1.6) | 1 (0.6) | |
| TTT treatment (%) | |||
| Yes | 15 (12.5) | 5 (3.1) | <0.01 |
| No | 110 (87.5) | 155 (96.9) | |
| Relative distance to the fovea (%) | 0.45 | ||
| 0 | 16 (12.8) | 16 (10.0) | |
| 0< x <1 | 9 (7.2) | 19 (11.9) | |
| 1–3 | 28 (22.4) | 52 (32.5) | |
| >3 | 35 (28.9) | 71 (44.4) | |
| Unknown | 37 (29.6) | 2 (1.3) | |
| Relative distance to optic nerve | 0.046 | ||
| 0 | 13 (10.4) | 13 (8.1) | |
| 0< x <1 | 10 (8.0) | 11 (6.9) | |
| 1–3 | 32 (25.6) | 59 (36.9) | |
| >3 | 33 (26.4) | 75 (46.9) | |
| Unknown | 37 (29.6) 2 (1.3) | 2 (1.3) | |
| Retinal detachment (%) | 0.004 | ||
| Yes | 22 (17.6) | 52 (32.5) | |
| No | 47 (37.6) | 69 (43.1) | |
| Not recorded | 56 (44.8) | 39 (24.4) | |
| Median delivered dose (Gy) | 85.3 (range: 80.9–100.8) | 84.6 (range: 80.7–88.5) | 0.047 |
| Median prescription point (mm) | 5 (range: 2.8–13.5) | 5 (range: 3.6–10.4) | <0.01 |
| Median plaque size (mm) | 18 (range: 14–22) | 18 (range: 14–24) | 0.35 |
OS = oculus sinister (left eye); OD = oculus dexter (right eye); CF = counting fingers; HM = hand motion; LP = light perception; TTT = transpupillary thermotherapy.
Bold p values denote <0.05.
Overall survival
Three-year OS was higher for the 103Pd group in univariate analysis: 95.8% vs. 90.2%, p = 0.047 (Fig. 1). Larger T stage (T3/T4), tumor height > 5 mm, and age > 60 years old were significant for worse OS on univariate. On MVA, only tumor height demonstrated a trend toward significance (hazard ratio [HR]: 1.99, 95% confidence interval [CI]: 0.91–4.38, p = 0.086). Radionuclide did not independently predict for OS (p = 0.889).
Fig. 1.
Estimated overall survival for treatment with 125I compared with 103Pd. Curve is truncated at 84 months. CI = confidence interval.
Disease control
For the entire population, 18 (6.3%) patients underwent secondary enucleation: 2 (11.1%) were due to pain without evidence of progression, 15 (83.3%) were secondary to local progression alone, and 1 (5.6%) was due to regional progression with pain. Pathology reports from enucleation are only available starting in 2009; for these three cases, all were confirmed on pathology specimen to have viable melanoma. The 5-year rate of secondary enucleation-free survival was statistically lower for the 125I cohort relative to the 103Pd cohort in univariate analysis (93.0% vs. 99.3%, p = 0.010) (Fig. 2a). MVA analysis revealed tumor height >5 mm (HR: 4.54, 95% CI: 1.25–16.42, p = 0.021) as the only significant prognostic factors. Radionuclide was not a statistically significant independent predictor for secondary enucleation (p = 0.170).
Fig. 2.
Estimated secondary enucleation (a) and distant metastases-free survival (b) for treatment with 125I compared with 103Pd. Curve is truncated at 84 months.
At 3 years, rate of distant metastases-free survival did not differ between the 125I and 103Pd cohorts in univariate analysis (94.1% vs. 97.1%, p = 0.254) (Fig. 2b). MVA analysis revealed that tumor height >5 mm stage was the only statistically significant factor to predict for distant metastases (HR: 6.61, 95% CI: 1.81–24.09, p = 0.004); radionuclide did not independently predict for distant metastases (p = 0.336).
VA and MVA
At baseline, 27 patients had severe visual impairment, defined as VA worse than 20/200. The 125I cohort had more patients with a severe visual impairment (16.0% vs. 4.4%, p = 0.002), but with no difference in the proportion with baseline VA better than 20/40 (42.4% vs. 51.9%, p = 0.11). At 3 years, all 27 patients continued to have vision worse than 20/200. As a result, this poor baseline VA subset was excluded for all analyses of posttreatment VA, leaving 258 eligible patients. Of these 258 patients, 152 (59%) had VA data available at 3 years.
At baseline, 82 103Pd patients and 53 125I patients had VA better than 20/40. At 3 years, the 103Pd cohort (52/82, 63%) had a higher percentage to maintain VA better than 20/40 than the 125I cohort (19/52, 35%) (p = 0.01). Table 2 details the factors included on MVA and their statistical significance for predicting for VA outcomes at 1 and 3 years. For predicting VA 20/40 or worse, MVA demonstrated 103Pd radionuclide to be a significant predictor at 1 (p = 0.017) and 3 years (p = 0.028). Tumor height < 5 mm had a either a strong trend toward significance at 1 year (p = 0.088) and was significant at 3 years (p = 0.017).
Table 2.
Multivariate analysis for visual outcomes at 1 and 3 years
| Visual outcomes | Variable | OR* | 95% CI | p-value |
|---|---|---|---|---|
| 1 years | ||||
| Preservation | Isotope: 103Pd vs. 125I | 2.70 | 1.20–6.25 | 0.02 |
| Tumor height: ≤5 mm vs. >5 mm | 2.04 | 0.90–4.55 | 0.09 | |
| Change in vision | Isotope: 103Pd vs. 125I | 1.44 | 0.41–5.26 | 0.55 |
| Tumor height: >5 mm vs. ≤5 mm | 0.56 | 0.16–1.89 | 0.35 | |
| Prior TTT: yes vs. no | 0.45 | 0.08–2.49 | 0.36 | |
| Loss | 103Pd vs. 125I | 1.45 | 0.66–3.57 | 0.32 |
| Tumor height: >5 mm vs. ≤5 mm | 0.59 | 0.26–1.35 | 0.21 | |
| Prior TTT: yes vs. no | 0.16 | 0.03–0.84 | 0.03 | |
| Distance from optic nerve: >1 vs. 0–1 | 0.64 | 0.27–1.51 | 0.31 | |
| 3 years | ||||
| Preservation | Isotope: 103Pd vs. 125I | 2.13 | 1.09–4.17 | 0.03 |
| Tumor height: ≤5 mm vs. >5 mm | 2.78 | 1.20–6.25 | 0.02 | |
| Notched plaque | 2.50 | 0.97–6.25 | 0.06 | |
| Change in vision | Isotope: 103Pd vs. 125I | 1.72 | 0.78–3.85 | 0.18 |
| Tumor height: >5 mm vs. ≤5 mm | 0.47 | 0.19–1.15 | 0.10 | |
| Prior TTT: yes vs. no | 0.42 | 0.11–1.62 | 0.21 | |
| Distance from optic nerve: >1 vs. 0–1 | 3.78 | 1.03–13.84 | 0.05 | |
| Loss | 103Pd vs. 125I | 1.45 | 0.65–3.23 | 0.35 |
| Tumor height: >5 mm vs. ≤5 mm | 0.28 | 0.12–0.64 | 0.00 | |
| Prior TTT: yes vs. no | 0.85 | 0.30–2.43 | 0.77 | |
| Distance from optic nerve: >1 vs. 0–1 | 3.40 | 1.27–9.11 | 0.02 |
TTT = transpupillary thermotherapy; CI = confidence interval;
OR = odds ratio.
Bold p value denotes <0.05.
Preservation: posttreatment acuity better than 20/40.
Loss: posttreatment acuity worse than 20/200.
1–2 degrees of change: posttreatment acuity improved, stable, or decreased by up to 1–2 lines on Snellen chart.
Seventy-one 125I patients and 111 103Pd patients had baseline vision better than or equal 20/200. At 3 years, the 125I cohort had a trend toward developing higher rates of VA 20/200 or worse at 3-year followup (33.8% [n = 24] vs. 22.5% [n = 25], p = 0.09). MVA for VA loss demonstrated tumor height >5 mm as the only significant factor at 1 (p = 0.025) and 3 years (p = 0.002) (Table 2). Radionuclide did not independently predict for development of severe visual impairment at 1 or 3 years.
In comparing the degree of visual changes, at 3-year, the 103Pd cohort had higher rates of improved vision, no change, or 1–2 lines of vision loss (65.5% vs. 52.5%) and lower rates of ≥6 lines of lost vision (10.8% vs. 27.5%, p = 0.03). MVA analysis for VA change of up to 1–2 lines of visual loss demonstrated no significant predictors (Table 2). Radionuclide was not significant on MVA for predicting degree of VA change at 3 years (p = 0.5, respectively).
VA and propensity score
MVA suggests that 103Pd radionuclide is a predictor for vision better than 20/40 at 1 and 3 years. To compare VA among more similarly identified patients, we used propensity score balancing and analysis. Twenty-three 125I patients were matched with 107 103Pd patients. Table 3 shows that these groups are well balanced in regard to gender, tumor height, diameter, T stage, baseline vision <20/40, and relative optic nerve distance. The 103Pd did have higher rates of retinal detachment (73.8% vs. 17.0%), whereas 125I patients continued to have greater relative proximity to the fovea (34.5% vs. 10.7%, p = 0.006). Within this matched cohort, 103Pd continues to be a significant predictor for VA preservation at 1 (p = 0.042) and 3 years (p = 0.014). Radionuclide was not a significant predictor for VA loss or changes in line vision.
Table 3.
Propensity score analysis, balance check
| Covariate | Reference level | 125I N = 23 (%) | 103Pd N = 107 (%) | Standardized difference |
|---|---|---|---|---|
| Maximum base diameter | <16 mm | 87.8 | 97.9 | 0.399 |
| T stage | T1/2 | 83.5 | 91.6 | 0.246 |
| Age | <60 | 26.1 | 54.9 | 0.612 |
| Tumor height | <5 mm | 78.1 | 92.3 | 0.410 |
| Gender | Male | 38.9 | 48.5 | 0.195 |
| Relative distance from optic nerve to tumor | Ratio ≤1 | 24.12 | 13.6 | −0.272 |
| Relative distance from fovea to tumor | Ratio ≤1 | 34.5 | 10.7 | −0.590 |
| Baseline vision | <20/40 | 48.7 | 57.7 | 0.181 |
| Retinal detachment | Yes | 17.0 | 73.8 | −1.386 |
Radiation retinopathy
Toxicity at 3 years or greater was evaluable in 195 patients (60.8%)—112 (57.4%) treated with Pd103 and 83 (43.6%) treated with I125—with available information. Overall, 73 patients developed radiation retinopathy. Salvage treatment for toxicity included laser photocoagulation (n = 15), bevacizumab (n = 15), and vitrectomy (n = 7).
At 3 years, radiation retinopathy and vitreous hemorrhage occurred less in the 103Pd cohort: 47.3% vs. 63.1%, p < 0.001 (Fig. 3). MVA demonstrated that 103Pd radionuclide was significantly associated with lower risk of radiation toxicity (HR: 1.88, 95% CI: 1.02–3.47, p = 0.043). Tumor height >5 mm (p = 0.968), TTT (p = 0.519), and distance from the optic nerve to tumor (p = 0.810) did not predict for higher risk of radiation retinopathy. No differences in rates of salvage treatments were seen between cohorts.
Fig. 3.
Estimated rate of developing radiation retinopathy or vitreous hemorrhage with 125I compared with 103Pd. Curve is truncated at 60 months.
Discussion
Since the COMS study, 125I EPB has been incorporated as a treatment option for potential eye/vision preservation for small and medium choroidal melanoma patients. However, despite 70% of patients having baseline VA 20/40 or better, only 34% of patients (49% of the 20/40 cohort) maintained 20/40 vision or better at 3 years (4). This demonstrates that there is substantial room for further improvement in post-EPB VA while maintaining oncologic outcomes.
To directly compare 125I and 103Pd, we investigated the VA of 258 patients with choroidal melanoma and baseline vision better than or equal to 20/200 treated with EPB at our single institution. We demonstrated that 103Pd cohort was associated with higher rates of visual preservation (defined as better than 20/40) at 3 years (39.6% vs. 19.1%, p = 0.006), with radionuclide remaining significant in adjusted MVA. Furthermore, within a more balanced cohort of patients—as created by propensity score analyses—103Pd continued to predict for improved VA preservation at 1 and 3 years. We also demonstrated that 103Pd was associated with a significantly lower risk of retinopathy at 3 years (39.6% vs. 71.6%, p < 0.001), with 103Pd radionuclide remaining a strong predictor of reduced toxicity on adjusted MVA. Though radionuclide independently and significantly predicted for VA preservation, it was not associated with degree of VA change or loss. Finally, the improved visual outcomes with 103Pd did not come at a detriment to oncologic outcomes: on adjusted analyses, 103Pd was not a predictor for secondary enucleation, distant metastases, or OS.
103Pd, due to its lower energy resulting in quicker dose fall-off, has been demonstrated dosimetrically to decrease dose to the macula and optic nerve relative to the 125I (7). Finger et al. subsequently investigated if this physical advantage translated into a clinical benefit. At 3 years, only 17.6% of 103Pd patients developed vision worse than 20/200 (19), compared with 43% of 125I patients at 3 years on the COMS medium study (4). The authors concluded 103Pd likely has higher rates of visual preservation than 125I; there are, however, many potential issues with this comparison including differences in baseline VA and patient comorbidities, tumor characteristics, and comparison of a retrospective cohort to an intention-to-treat phase III trial. Our study builds on the findings by Finger et al. by directly comparing patients treated with 103Pd to 125I. We demonstrate that 103Pd is associated with improved VA and decreased radiation retinopathy; these results, however, are probably due in part to selection bias and difference in patient characteristics. However, these finding, especially in light of both MVA and propensity score balancing, is provocative and suggests that choice of radionuclide may help better maintain vision <20/40. Though radionuclide independently and significantly predicted for VA preservation, it was not associated with degree of VA change or VA loss (defined as VA worse than 20/200).
Limitations of this study include its retrospective design, potential for selection bias due to the nonrandomized treatment cohorts, difference in time periods of patient treatment (with 103Pd use starting in 2004), and differences in baseline patient and tumor characteristics and median followup periods between cohorts. Moreover, although we accounted for numerous variables known to be predictive for visual outcomes after EPB, many patients had missing data; these unknown missing values—higher in the iodine cohort—for retinal detachment and distance to optic nerve and fovea, likely further contributed to our observed VA differences. We did attempt to minimize difference in baseline prognostic differences using two separate statistical models: multivariable analyses and propensity score balancing; however, our statistical measures cannot completely overcome these important differences. Strengths of this study include its relatively large patient numbers, homogenous patient treatment methods and followup/surveillance schedule, and well-documented patient, tumor, and treatment characteristics for use in MVA to adjust for potential confounding variables. To our knowledge, we also are the first to report on oncologic and VA outcomes directly comparing, albeit retrospective cohort study, 103Pd with 125I EPB.
In conclusion, treatment with either 125I or 103Pd results in excellent tumor control and low rates of secondary enucleation. However, 103Pd radionuclide use was associated with increased probability of long-term VA preservation and reduced risk of radiation retinopathy compared with 125I. Additionally, tumor height is a strong significant predictor of VA outcomes, toxicity outcomes, and local control probability, most likely due to both tumor bulk and radiation dosimetry factors. As this study is subject to limitations common to all retrospective reviews, a prospective clinical trial comparing EPB radionuclides is warranted.
Footnotes
Conflict of interest: None.
References
- 1.Emara K, Weisbrod DJ, Sahgal A, et al. Stereotactic radiotherapy in the treatment of juxtapapillary choroidal melanoma: preliminary results. Int J Radiat Oncol Biol Phys. 2004;59:94–100. doi: 10.1016/j.ijrobp.2003.10.007. [DOI] [PubMed] [Google Scholar]
- 2.Collaborative Ocular Melanoma Study G. The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma: V. Twelve-year mortality rates and prognostic factors: COMS report No. 28. Arch Ophthalmol. 2006;124:1684–1693. doi: 10.1001/archopht.124.12.1684. [DOI] [PubMed] [Google Scholar]
- 3.Diener-West M, Reynolds SM, Agugliaro DJ, et al. Development of metastatic disease after enrollment in the COMS trials for treatment of choroidal melanoma: Collaborative Ocular Melanoma Study Group Report No. 26. Arch Ophthalmol. 2005;123:1639–1643. doi: 10.1001/archopht.123.12.1639. [DOI] [PubMed] [Google Scholar]
- 4.Melia BM, Abramson DH, Albert DM, et al. Collaborative ocular melanoma study (COMS) randomized trial of I-125 brachytherapy for medium choroidal melanoma. I. Visual acuity after 3 years COMS report no. 16. Ophthalmology. 2001;108:348–366. doi: 10.1016/s0161-6420(00)00526-1. [DOI] [PubMed] [Google Scholar]
- 5.Nag S, Quivey JM, Earle JD, et al. The American Brachytherapy Society recommendations for brachytherapy of uveal melanomas. Int J Radiat Oncol Biol Phys. 2003;56:544–555. doi: 10.1016/s0360-3016(03)00006-3. [DOI] [PubMed] [Google Scholar]
- 6.Wen JC, Oliver SC, McCannel TA. Ocular complications following I-125 brachytherapy for choroidal melanoma. Eye. 2009;23:1254–1268. doi: 10.1038/eye.2009.43. [DOI] [PubMed] [Google Scholar]
- 7.Rivard MJ, Coursey BM, DeWerd LA, et al. Update of AAPM Task Group No. 43 Report: a revised AAPM protocol for brachytherapy dose calculations. Med Phys. 2004;31:633–674. doi: 10.1118/1.1646040. [DOI] [PubMed] [Google Scholar]
- 8.Finger PT, Lu D, Buffa A, et al. Palladium-103 versus iodine-125 for ophthalmic plaque radiotherapy. Int J Radiat Oncol Biol Phys. 1993;27:849–854. doi: 10.1016/0360-3016(93)90459-9. [DOI] [PubMed] [Google Scholar]
- 9.Nath R, Anderson LL, Luxton G, et al. Dosimetry of interstitial brachytherapy sources: recommendations of the AAPM radiation therapy Committee Task group No. 43 American association of Physicists in Medicine. Med Phys. 1995;22:209–234. doi: 10.1118/1.597458. [DOI] [PubMed] [Google Scholar]
- 10.Ray SK, Bhatnagar R, Hartsell WF, et al. Review of eye plaque dosimetry based on AAPM Task group 43 recommendations American association of Physicists in Medicine. Int J Radiat Oncol Biol Phys. 1998;41:701–706. doi: 10.1016/s0360-3016(97)00568-3. [DOI] [PubMed] [Google Scholar]
- 11.Thomson RM, Furutani KM, Pulido JS, et al. Modified COMS plaques for 125I and 103Pd iris melanoma brachytherapy. Int J Radiat Oncol Biol Phys. 2010;78:1261–1269. doi: 10.1016/j.ijrobp.2009.12.002. [DOI] [PubMed] [Google Scholar]
- 12.Thomson RM, Taylor RE, Rogers DW. Monte Carlo dosimetry for 125I and 103Pd eye plaque brachytherapy. Med Phys. 2008;35:5530–5543. doi: 10.1118/1.3002412. [DOI] [PubMed] [Google Scholar]
- 13.American Brachytherapy Society – Ophthalmic Oncology Task Force. Electronic address pec, Committee AO. The American Brachytherapy Society consensus guidelines for plaque brachytherapy of uveal melanoma and retinoblastoma. Brachytherapy. 2014;13:1–14. doi: 10.1016/j.brachy.2013.11.008. [DOI] [PubMed] [Google Scholar]
- 14.Chiu-Tsao ST, Astrahan MA, Finger PT, et al. Dosimetry of (125)I and (103)Pd COMS eye plaques for intraocular tumors: report of Task Group 129 by the AAPM and ABS. Med Phys. 2012;39:6161–6184. doi: 10.1118/1.4749933. [DOI] [PubMed] [Google Scholar]
- 15.Finger PT, Zhou D, Kalach N, et al. 103Pd versus 125I ophthalmic plaque brachytherapy: preoperative comparative radiation dosimetry for 319 uveal melanomas. J Radiat Oncol. 2014;3:409–416. doi: 10.1007/s13566-014-0149-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Finger PT, Berson A, Ng T, et al. Palladium-103 plaque radiotherapy for choroidal melanoma: an 11-year study. Int J Radiat Oncol Biol Phys. 2002;54:1438–1445. doi: 10.1016/s0360-3016(02)03751-3. [DOI] [PubMed] [Google Scholar]
- 17.Schulze-Bonsel K, Feltgen N, Burau H, et al. Visual acuities “hand motion” and “counting fingers” can be quantified with the freiburg visual acuity test. Invest Ophthalmol Vis Sci. 2006;47:1236–1240. doi: 10.1167/iovs.05-0981. [DOI] [PubMed] [Google Scholar]
- 18.Holladay JT. Proper method for calculating average visual acuity. J refractive Surg. 1997;13:388–391. doi: 10.3928/1081-597X-19970701-16. [DOI] [PubMed] [Google Scholar]
- 19.Finger PT, Chin KJ, Duvall G, Palladium-103 for Choroidal Melanoma Study G Palladium-103 ophthalmic plaque radiation therapy for choroidal melanoma: 400 treated patients. Ophthalmology. 2009;116:790–796. 796.e1. doi: 10.1016/j.ophtha.2008.12.027. [DOI] [PubMed] [Google Scholar]